A Residential Builder’s Guide to Steel Beam Span Tables — When You Can Estimate from the Manual, and When You Need an Engineer

Two construction workers in hard hats reviewing building plans and a tablet on a construction site

Every residential builder doing extension work in Victoria has used a span table. The OneSteel Hot Rolled and Structural Steel Products manual, the BlueScope DuraGal design guide, the ASI residential steel application charts — each one comes with span charts that turn “can a 200UB22 carry the roof and first-floor over a 4.2 m opening” or “does a 250UB31 need a mid-span column over 5.5 m” into a 90-second estimate.

But a span table is not a structural design. It is a budgeting tool, a preliminary sizing tool, and a tool for the conversation with the engineer that follows — not a substitute for the engineer. Builders who treat span tables as a stand-in for engineering sign-off eventually meet trouble at permit review, at insurance time, or at the on-site moment when an actual load lands somewhere the table did not anticipate. This article draws the boundary that often gets blurred in everyday practice — where a builder can responsibly use a span table, and where the drawings have to be sent to an RPEng-registered structural engineer for calculation and sign-off. The audience is residential builders, draftspersons and owner-builders in Victoria and New South Wales doing single-storey extensions, open-plan renovations and load-bearing wall removals.

What a span table actually gives you

To understand the boundary, it helps to understand what is sitting behind the table.

Every residential steel beam span table is built on a fixed set of assumptions: a particular section (say 200UB22), a steel grade (typically 300PLUS or 350), a single simply-supported span with no continuity, a defined load combination (such as “single-storey roof plus one floor with standard imposed load”), and a target deflection limit (commonly span/250 or span/300). Within those assumptions, the maximum allowable span on the chart is valid.

For a single-storey extension where you are putting a single 4 m beam over an internal opening, carrying roof truss loading from above and no unusual point loads or cantilevers, the table’s preliminary sizing is genuinely useful. The builder can put that estimate into the customer quote, the material pre-order, and the first conversation with the engineer. It saves time.

Span tables also help builders sense-check proportion. If a client wants to span a 6 m first-floor opening on a 150 PFC, the chart tells you in one glance that the combination is not real — and “filtering out the impossible early” is one of the most useful things a span table does.

What a span table does not give you

What the table delivers is the answer to the question the chart assumes you are asking. The moment your project drifts past those assumptions, the table goes silent and the work has to move to a structural engineer.

The first boundary is continuity. Span tables are simply-supported single-span. The moment a beam is continuous over two or more spans — for example, an 8 m beam landing on a mid-column and behaving as two 4 m sections — the distribution of bending moments at midspan and over the support changes completely. The simply-supported value is conservative for span moments but actively misleading for support reactions, which can mean the column or footing underneath is incorrectly sized.

The second boundary is cantilevers. The chart values are for beams supported at both ends. A cantilever bending moment behaves differently — a 1.5 m cantilevered balcony beam can require a heavier section than a 6 m simply-supported beam, because the root moment is not balanced by support at the far end. Cantilevers always go to an engineer.

The third boundary is point loads. Span tables assume uniformly distributed loading. The moment a beam carries a 200 kg air-conditioning unit, takes the load of a column landing from the floor above, or receives a secondary beam transferring concentrated load from another bay, the loading profile is no longer what the chart calculated.

The fourth boundary is non-standard or high imposed loads. Imposed load is not one number. Residential floors, balconies, garages, roof gardens, storage rooms, plant rooms, solar array zones — each carries a different prescribed value under AS/NZS 1170. Standard residential floor sits at 1.5 kPa; balconies at 2.0 kPa; roof gardens and storage areas climb to 3.0 kPa or higher. A span table assumes a “default residential combination” and stops being usable as soon as the project departs from it.

A woman in a suit reviews architectural plans with a pen while a computer screen displays a 3D model nearby.

The fifth boundary is lateral stability. A tall, slender UB section without floor restraint above will fail by lateral-torsional buckling well before it reaches its calculated bending capacity, and the chart’s allowable span has to be drastically reduced. The “fully laterally restrained” assumption baked into most charts fails in many practical residential cases — for example, an exposed beam running below a living room ceiling with no slab or truss bracing it laterally must be re-evaluated as unrestrained.

The sixth boundary is connections. How two beams join, how a beam sits onto a brick pier, how it connects down to timber trusses, whether a baseplate and anchor bolts are required — these belong to AS 4100 Section 9 and span tables touch none of them.

The seventh boundary is fire resistance. Multi-unit residential and commercial projects require structural steel to carry an FRL under the NCC, which can involve intumescent paint, fire boards or section upsizing. Span tables do not consider FRL at all.

The eighth boundary is construction categoryCC1, CC2, CC3. AS/NZS 5131 sorts steelwork by consequence class, with most single-storey residential work in CC2 and multi-unit work in CC3. The CC class drives fabricator qualification, welder certification, inspection frequency and documentation requirements. None of that is something a builder selects from a span table. It is something the engineer determines under AS 4100, AS/NZS 1170 and AS/NZS 5131, and writes into the design notes.

A practical decision flow

Translated into a checklist a builder can actually use on site or at budgeting time, the boundary works like this.

First, ask whether the beam is single-span, simply-supported, uniformly distributed loaded, laterally restrained by the floor above, and inside the default single-storey residential load combination. If all five hold and the span sits in the 3 to 5.5 m residential range, the span table’s estimate is appropriate for budgeting and for the early engineer conversation.

Second, anything that involves removing a load-bearing internal wall in an open-plan renovation, converting an external wall to a full-height glass slider, exposing a 6 m or longer beam below a ceiling, spanning a first-floor opening greater than 4.5 m, cantilevering a balcony beyond 1.2 m, or carrying a roof garden or large solar array — sends the drawings to an engineer.

Third, anything that involves multi-span continuity, point loads, cantilevers, lateral-torsional buckling risk, baseplate design or anchor bolt design — engineer, and RPEng-registered.

Fourth, on CC2 work the structural drawings carry the engineer’s signature and stamp; on CC3 work, in addition to design sign-off, fabrication and welding follow the AS/NZS 5131 CC3 process, including non-destructive testing and process inspection records.

Fifth, the engineer’s drawings go to the fabricator’s in-house CAD team for shop drawings. Shop drawings are not the engineer’s deliverable — the fabricator produces them and the engineer reviews and signs them back. This step translates “design intent” into the cut-list, weld-list and assembly sequence the workshop can actually execute.

Why a builder’s own table-reading rarely saves money

Owner-builders and less-experienced small builders sometimes treat “I’ve done a dozen of these, I can just pick the section off the chart myself” as a reasonable shortcut. The cost layers behind that view are worth being clear about.

The first is compliance cost. Victoria’s building permit process requires a structural engineer’s design support for Class 1a residential work above a certain area or involving structural alterations. If a builder picks the section from the chart, the surveyor will require an engineer’s calc anyway. If the calc comes back and the picked section is wrong, the project either re-drafts, replaces the beam, or argues for retention of the chosen section — none of which is cheap.

The second is insurance cost. Most builders’ home warranty insurance and public liability cover require structural elements to be designed by a qualified engineer. If a builder-picked beam ever shows signs of distress — settlement, cracking, connection movement — insurers commonly cite “not engineer-designed” as grounds for declining cover.

The third is rework cost. Once a structural steel beam is cut, welded and installed into the wall, discovering it is undersized is rarely a quick fix. Removing an installed lintel and replacing it pulls in secondary beams, floor framing, roof framing and surrounding wall work — typically 20 to 50 times the cost of the engineer’s original fee.

The fourth is liability cost. Building industry liability does not evaporate in five years. A 15-year-old steel beam that develops capacity issues triggers an investigation whose first question is whether the original work carried engineer sign-off. Builders without that sign-off carry personal liability.

A workable collaboration model

The builders we work with most often are small-to-mid residential outfits doing extensions and major renovations in the AUD 150,000 to 2.5 million range. At that scale, projects do not usually carry a dedicated project manager — structural steel coordination falls to the site supervisor or to the builder personally. A smooth collaboration model usually looks like this.

At the budgeting stage, the builder uses a span table to set a preliminary section range and quotes the client a “structural steel including fabrication and installation in the range AUD X to Y” envelope. The client is told explicitly that the engineer’s final section sizing may move within that range.

At the design stage, the architect’s or draftsperson’s preliminary drawings go to an RPEng-registered structural engineer, who produces structural drawings and calculations. This usually takes one to three weeks; complex jobs longer. When the builder and the engineer have an established working relationship, that compresses to within a week.

At the quoting stage, the engineer’s structural drawings go to the fabricator for shop drawings and a firm quote. Our residential structural steel service and our steel beam fabrication and installation service typically turn structural drawings into shop drawings and a binding quote inside 5 to 10 working days.

At the fabrication and installation stage, the work follows the AS/NZS 5131 process for the project’s CC class — fit-up inspection, weld inspection, transport, on-site rigging. Most residential work is CC2; shop drawings go back to the engineer for review, welding is performed by certified welders, and the fabricator issues a Certificate of Compliance on completion.

The boundary works in the builder’s favour

A lot of builders treat “needs an engineer’s sign-off” as an extra cost. Read the other way, it is a protection. Once the engineer signs the drawing, structural responsibility sits with the engineer, and the builder’s role moves back inside the clean limits of “build to the drawings, fabricate to the standard, deliver against the contract”. A builder who does everything themselves carries design, build, compliance and insurance risk in one package; a builder who knows what to send to the engineer, what to estimate themselves, and what to send to the fabricator runs projects faster, more reliably and more cheaply.

Span tables are useful tools for builders, provided they are treated as estimating tools rather than as design documents. AS 4100 and AS/NZS 5131 assign responsibility along the chain from mill to handover, and the cost of crossing that chain casually is much higher than the cost of working with it. The structural role of steel in residential work has grown noticeably in recent years — bigger openings, more exposed expression, more complex open-plan layouts — and the boundary matters more in that environment, not less.

Span table conventions, CC class assignments, RPEng sign-off and Victorian building permit processes in this article reflect current practice in the Australian residential structural steel market; specific section choice, connection design and compliance requirements must be confirmed by the project’s RPEng-registered structural engineer and against the latest editions of the relevant standards published by the ABCB, Standards Australia and Engineers Australia. This article is not engineering advice and does not replace design or sign-off by an RPEng-registered structural engineer.

Leave a Reply

Your email address will not be published. Required fields are marked *